BitCrack - A tool for brute-forcing private keys

[holy_ship]

Ever since 2015, we are stuck at puzzle 66, you should double the key space size each bit you go up.

You're mistaken. P64 was depleted only 1 year ago.

P65 was depleted by different algo.

So, we are not really STUCK with 66

[digaran]

As a developer of 30 years I decided to fully understand SHA256 I would create a valid sha256 hash using binary only and then see how far I could reverse each step.

What were you developing in 30 years? If it was anything crypto/ security related, you should contact your clients and tell them they are compromised. 😅
Kidding.
Welcome on board!
I can't really understand your goal here, why are you trying to reverse a sha256 of a hex private key, which is BTW 17 characters and it's hash has no use in at least in Bitcoin.
Private keys are irrelevant to hashes and addresses, once you generate your priv key, then derive your public key from it, after that you perform sha256 on the public key and then rmd160 on that sha256 hash, it's not over yet,  When you have rmd160 you'd add version/network byte then use a double sha256 to derive the checksum in order to have an address.

So if you want to reverse anything, try by reversing the rmd160 hash of puzzle #66, when you find which sha256 hash outputs the rmd160, you have either found a collision or the original hash, but knowing that is impossible, anyways after having the sha256 hash, you'd need to reverse it back to the public key, and guess what?

There is no need to reverse 2 hash functions to find the public key, you already have access to public keys of puzzle 130, 135 etc which are housing more coins than 66.
I suggest you focus on public keys if you are good with math.😉

[Zipp_1199]

Hi guys,

I was reading this thread and found it interesting. Just a few questions get in my mind:
What if I find a zero-bit privkey address? How long would it take me to find a private key for this address, knowing only the address? Is it a million years, assuming that I don't have a public key?

[citb0in]

Hi guys,

I was reading this thread and found it interesting. Just a few questions get in my mind:
What if I find a zero-bit privkey address? How long would it take me to find a private key for this address, knowing only the address? Is it a million years, assuming that I don't have a public key?

it will take you 1.42 million years to find the solution to your zero-bit question

[digaran]

Hi guys,

I was reading this thread and found it interesting. Just a few questions get in my mind:
What if I find a zero-bit privkey address? How long would it take me to find a private key for this address, knowing only the address? Is it a million years, assuming that I don't have a public key?

Can you show us what you mean by zero bit private key? If it has "zero" bits then how can you derive an address from it? But there is a solution for you, it's not related to zero or 256, you can find 2^96 different private keys for every single address out there, under one condition, you'd have to brute force 2^80 keys to find one collision.(in theory)
Without public key, only having the address and knowing the exact range, you can go ahead and try puzzle #66, even if you halve the range, you'd still need to grind 2^65 keys.😉

[Zipp_1199]

Hi guys,

I was reading this thread and found it interesting. Just a few questions get in my mind:
What if I find a zero-bit privkey address? How long would it take me to find a private key for this address, knowing only the address? Is it a million years, assuming that I don't have a public key?

Can you show us what you mean by zero bit private key? If it has "zero" bits then how can you derive an address from it? But there is a solution for you, it's not related to zero or 256, you can find 2^96 different private keys for every single address out there, under one condition, you'd have to brute force 2^80 keys to find one collision.(in theory)
Without public key, only having the address and knowing the exact range, you can go ahead and try puzzle #66, even if you halve the range, you'd still need to grind 2^65 keys.😉

Sorry for confusion. I just try to understand the whole thing, that is why i am asking. For instance a wallet 1BgGZ9tcN4rm9KBzDn7KprQz87SZ26SAMH had a very simple private key of 0x0000000000000000000000000000000000000000000000000000000000000001? This means it is easy to crack it as the only thing is needed is checksum?
So if i would ever find an address with a privkey of 0x0000000000000000000000000000000000000000000000000000000000000001 but no public key - is it something i would have to run 2^65 keys to open it?

[citb0in]

Start HERE

[MrFreeDragon]

I remember I saw somewhere the benchmark table with the speeds for BitCrack with different nvidia cards (2080,3080,4080, .... ti versions, Tesla, etc). Can you please share it or write the current benchmarks (actual in 2024 with the most common hardware).

[zahid888]

I remember I saw somewhere the benchmark table with the speeds for BitCrack with different nvidia cards (2080,3080,4080, .... ti versions, Tesla, etc). Can you please share it or write the current benchmarks (actual in 2024 with the most common hardware).

https://bitcointalk.org/index.php?topic=4453897.msg55552855#msg55552855

[NotATether]

I remember I saw somewhere the benchmark table with the speeds for BitCrack with different nvidia cards (2080,3080,4080, .... ti versions, Tesla, etc). Can you please share it or write the current benchmarks (actual in 2024 with the most common hardware).

https://bitcointalk.org/index.php?topic=4453897.msg55552855#msg55552855

Unrelated, but I am just wondering have you ever found any keys from all those sub-ranges you were searching in for on Bitcrack for a long time, or did you not find anything?

[unamic]

What is the best Settings for an RTX 3090?
I have this for now -b 128 -t 256 -p 1024 but is this the best setting?

[qq569160818]

Can try different combinations multiple times, personal recommendation  256   512   1024  or   512    1024    2048？

[rexxx97]

Hello,
there is a benchmark database of nvidia rtx 30s and 40s?

[raschwarz]

Hello,
there is a benchmark database of nvidia rtx 30s and 40s?

hi, I am also looking for such table, but no luck till now

My results with Ubuntu 22.04, nVidia 545.x.x drivers and CUDA v12.3 on i5 CPU, 96 GB DDR5 RAM, single RTX 4090 (water cooled, from MSI)

Software	t	b	p	result	Watts
cuBitCrack	256	256	512	~2975 MKey/s	466
cuBitCrack	256	256	1024	~3040 MKey/s	467
cuBitCrack	128	512	1024	~3080 MKey/s	467
cuBitCrack	256	256	2048	~3080 MKey/s	467
cuBitCrack	128	512	2048	~3165 MKey/s	464
cuBitCrack	128	256	4096	~3130 MKey/s	464
cuBitCrack	256	128	4096	~3090 MKey/s	464

I don't know if it can be better. I played with param values longer time, but many time I was stopped by "[Info] Error: out of memory" or  "[Info] Error: too many resources requested for launch".

[WanderingPhilospher]

Hello,
there is a benchmark database of nvidia rtx 30s and 40s?

hi, I am also looking for such table, but no luck till now

My results with Ubuntu 22.04, nVidia 545.x.x drivers and CUDA v12.3 on i5 CPU, 96 GB DDR5 RAM, single RTX 4090 (water cooled, from MSI)

Software	t	b	p	result	Watts
cuBitCrack	256	256	512	~2975 MKey/s	466
cuBitCrack	256	256	1024	~3040 MKey/s	467
cuBitCrack	128	512	1024	~3080 MKey/s	467
cuBitCrack	256	256	2048	~3080 MKey/s	467
cuBitCrack	128	512	2048	~3165 MKey/s	464
cuBitCrack	128	256	4096	~3130 MKey/s	464
cuBitCrack	256	128	4096	~3090 MKey/s	464

I don't know if it can be better. I played with param values longer time, but many time I was stopped by "[Info] Error: out of memory" or  "[Info] Error: too many resources requested for launch".

I would suggest to stop using bitcrack until it is updated, if ever.

A 4090 using the vanitysearch with bitcrack option, gets 5000 MKey/s, and using KeyHuntCuda, a 4090 gets over 6000 MKey/s.

You are wasting MKey/s by still trying to use bitcrack, IMO.

[raschwarz]

Hello,
there is a benchmark database of nvidia rtx 30s and 40s?

hi, I am also looking for such table, but no luck till now

My results with Ubuntu 22.04, nVidia 545.x.x drivers and CUDA v12.3 on i5 CPU, 96 GB DDR5 RAM, single RTX 4090 (water cooled, from MSI)

Software	t	b	p	result	Watts
cuBitCrack	256	256	512	~2975 MKey/s	466
cuBitCrack	256	256	1024	~3040 MKey/s	467
cuBitCrack	128	512	1024	~3080 MKey/s	467
cuBitCrack	256	256	2048	~3080 MKey/s	467
cuBitCrack	128	512	2048	~3165 MKey/s	464
cuBitCrack	128	256	4096	~3130 MKey/s	464
cuBitCrack	256	128	4096	~3090 MKey/s	464

I don't know if it can be better. I played with param values longer time, but many time I was stopped by "[Info] Error: out of memory" or  "[Info] Error: too many resources requested for launch".

I would suggest to stop using bitcrack until it is updated, if ever.

A 4090 using the vanitysearch with bitcrack option, gets 5000 MKey/s, and using KeyHuntCuda, a 4090 gets over 6000 MKey/s.

You are wasting MKey/s by still trying to use bitcrack, IMO.

Thanks for reply, you are right. I am not using it anymore, just made test before and want to share some results.
I went to kangaroos/bsgs where is so much space for improvements and max 4090 utilization :-)

[Olek5k]

Hello, I also want to join the key finder (1000 BTC Bitcoin Challenge Transaction) but I don't know how to install BitCrack program (I have one rtx 2060 video card I want to try it), is it possible to have a detailed description or a video how to install this BitCrack program step by step and set it up correctly, I would like to try the selective method.

[coolindark]

Question:

Is it possible to make search specific spaces in the priv-key? Like shown below in X:

00000000000000000000000000000XXXXXXXX000000000000000000000000000

[iceland2k14]

Question:

Is it possible to make search specific spaces in the priv-key? Like shown below in X:

00000000000000000000000000000XXXXXXXX000000000000000000000000000

Yes this is done by using -stride option

[coolindark]

Question:

Is it possible to make search specific spaces in the priv-key? Like shown below in X:

00000000000000000000000000000XXXXXXXX000000000000000000000000000

Yes this is done by using -stride option

Yes, i got it but could not figure out how can i correctly mark the start-end points. Can you give me the stride setting for this example? This will help me.